The Epigenetic Factor BORIS/CTCFL Regulates the NOTCH3 Gene Expression in Cancer Cells

Biochimica et Biophysica Acta 1839 (2014) 813–825

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbagrm

The epigenetic factor BORIS/CTCFL regulates the NOTCH3 gene expression in cancer cells

Michele Zampieri a,b, Fabio Ciccarone a,b, Rocco Palermo c, Samantha Cialﬁ d, Claudio Passananti e, Sabina Chiaretti a, Daniela Nocchia a, Claudio Talora d, Isabella Screpanti b,d,⁎,PaolaCaiafaa,b,⁎⁎ a Department of Cellular Biotechnologies and Hematology, Faculty of Pharmacy & Medicine, Sapienza University of Rome, Rome, Italy b Pasteur Institute-Fondazione Cenci Bolognetti, Rome, Italy c Center for Life Nano Science@Sapienza, Istituto Italiano di Tecnologia, Rome, Italy d Department of Molecular Medicine, Faculty of Pharmacy & Medicine, Sapienza University of Rome, Rome, Italy e Institute of Molecular Biology & Pathology CNR, Department of Molecular Medicine, Sapienza University of Rome, Rome, Italy article info abstract

Article history: Aberrant upregulation of NOTCH3 gene plays a critical role in cancer pathogenesis. However, the underlying Received 24 January 2014 mechanisms are still unknown. We tested here the hypothesis that aberrant epigenetic modiﬁcations in the Received in revised form 23 June 2014 NOTCH3 promoter region might account for its upregulation in cancer cells. We compared DNA and histone Accepted 24 June 2014 methylation status of NOTCH3 promoter region in human normal blood cells and T cell acute lymphoblastic leu- Available online 28 June 2014 kemia (T-ALL) cell lines, differentially expressing NOTCH3. We found that histone methylation, rather than DNA Keywords: hypomethylation, contributes towards establishing an active chromatin status of NOTCH3 promoter in NOTCH3 T cell acute lymphoblastic leukemia overexpressing cancer cells. We discovered that the chromatin regulator protein BORIS/CTCFL plays an important Oncogene role in regulating NOTCH3 gene expression. We observed that BORIS is present in T-ALL cell lines as well as in cell Cancer testis antigen lines derived from several solid tumors overexpressing NOTCH3. Moreover, BORIS targets NOTCH3 promoter in Chromatin cancer cells and it is able to induce and to maintain a permissive/active chromatin conformation. Importantly, DNA/histone methylation the association between NOTCH3 overexpression and BORIS presence was conﬁrmed in primary T-ALL samples from patients at the onset of the disease. Overall, our results provide novel insights into the determinants of NOTCH3 overexpression in cancer cells, by revealing a key role for BORIS as the main mediator of transcriptional deregulation of NOTCH3. © 2014 Elsevier B.V. All rights reserved.

1. Introduction type, tissue context and the member of the Notch receptor family ac- tivated [3,4]. The human neurogenic locus notch homolog protein 3 (NOTCH3) An oncogenic role of NOTCH3 pathway was identified in T cell acute gene encodes the third discovered mammalian member of an evolu- lymphoblastic leukemia (T-ALL), since constitutive expression of the ac- tionary conserved family of transmembrane protein receptors (NOTCH tivated intracellular domain of NOTCH3 (NOTCH3-IC) has been shown 1–4). Notch receptors are strictly required for normal embryonic devel- to induce T-cell neoplasias in transgenic mice [5] and the overexpres- opment and for homeostasis of a variety of adult tissues by controlling sion of NOTCH3 in the vast majority of T-ALL validates this finding in the balance between proliferation versus differentiation of stem and/ humans [6]. or progenitor cells [1,2]. Aberrant NOTCH3 signaling has been also observed in several solid A deregulated Notch pathway has been found to be involved in tumors, where NOTCH3 activation favors cell proliferation, resistance the development of a wide variety of human tumors with either an to apoptotic stimuli, metastatic capability and maintenance of stem oncogenic or a tumor-suppressive role depending on the cancer cell-like features [7]. Moreover, a correlation between NOTCH3 expression and poor prog- fi ⁎ Correspondence to: I. Screpanti, Department of Molecular Medicine, Faculty of nostic outcome has been described in primary tumors thus con rming Pharmacy & Medicine, Sapienza University of Rome, Viale Regina Elena 291, 00161 NOTCH3 involvement in cancer pathogenesis [8–10]. Rome, Italy. Tel.: 0039(06)44700816. Molecular details of the genetic mechanisms (i.e. gene amplification, ⁎⁎ Correspondence to: P. Caiafa, Department of Cellular Biotechnologies and chromosome translocation) causing NOTCH3 receptor activation as well Hematology, Faculty of Pharmacy & Medicine, Sapienza University of Rome, Via Chieti 7, as the consequent activation of signaling are emerging in the context of 00161 Rome, Italy. Tel.: 0039(06)49976530. fi – E-mail addresses: [email protected] (I. Screpanti), [email protected] speci ctumors[11 15], while the causes of NOTCH3 gene overexpres- (P. Caiafa). sion in cancer cells remain poorly understood.

Epigenetic processes often contribute to the deregulated expression As with respect to PBMCs, the NOTCH3 silencing could be explained of genes critical to the development and progression of cancer. The fact by both the low H3K4me3 and high H3K27me3 amounts. However, this that the transcription start site (TSS) of NOTCH3 localizes within a CpG condition does not depend on the heterochromatization as indicated by island prompted us to investigate a possible correlation between the absence of H3K9me3, whose presence instead characterizes SKW3 NOTCH3 gene DNA methylation status and gene expression extending cells, which do not express NOTCH3. the analysis to histone methylation. We report here that the cancer testis antigen (CTA) BORIS/CTCFL 2.2. CTCF and BORIS differentially occupy NOTCH3 promoter in normal (Brother Of Regulator of Imprinted Sites)/CTCF-like protein), paralog blood cells and T-ALL cell lines of the widely expressed chromatin insulator CTCF (CCCTC-binding factor) [16,17], is involved in the epigenetic deregulation leading to To search for potential trans-acting protein factors responsible for NOTCH3 upregulation in cancer cells. In particular, BORIS presence mir- epigenetic regulation of NOTCH3 gene expression, the Genomatix rors NOTCH3 overexpression specifically in T-ALL established cell lines MatInspector software was used [18,19]. The inspection of as well as in primary T-ALL human samples from patients at the onset about 2000 bp of the NOTCH3 5-prime sequence (transcript ID of the disease. Furthermore, BORIS binding to NOTCH3 promoter is ENST00000263388) revealed the presence of eight regions located associated with histone modifications typical of permissive/active chro- in the proximity of the TSS (from − 391 to +259) which match matin conformation. Overall, our data highlight a key role played by with the CTCF binding sequence (data not shown). Three of them ex- BORIS in the abnormal upregulation of NOTCH3 in cancer cells. hibited a high degree of similarity to CTCF binding sites with a qual- ity rating above the algorithm defined cut off value of 0.8 (Fig. 2A, 2. Results sites 1, 2 and 3). Furthermore, same result was obtained by compar- ing the same region with previously published 20-bp binding motifs 2.1. NOTCH3 gene expression and promoter epigenetic profile in normal for CTCF [20] (Fig. 2B). This was also suggested by ChIP-seq data from blood cells and in T-ALL cell lines the ENCODE consortium accessed through the UCSC genome browser (http://genome.ucsc.edu) [21–23] (Supplementary Fig. S1). Whether or not DNA methylation of the NOTCH3 promoter restricts CTCF DNA-target sites could also be bound by CTCFL/BORIS, a CTCF- NOTCH3 expression was investigated. paralog protein that may have epigenetic CTCF-antagonistic regulative NOTCH3 mRNA level analysis was performed by real-time RT-PCR in functions on CTCF binding regions of target genes [17,24–26]. normal blood cells [peripheral blood mononuclear cells (PBMCs), To address whether CTCF and BORIS are both involved in the regula- CD4+ and CD8+ T-lymphocytes] and in T-ALL cell lines differentially tion of NOTCH3 expression, transcript levels of CTCF and BORIS were an- expressing NOTCH3 (MOLT3 and SKW3). Fig. 1A shows that the alyzed by RT-PCR in normal blood cells and T-ALL cell lines, to compare NOTCH3 mRNA level varied significantly with over 500-fold difference their expression with that of NOTCH3 (Fig. 1A). between the lowest and highest expressing cell types (MOLT3 vs. To detect all the 23 known mRNA variants of BORIS [27], its expres- CD8+ T-lymphocytes). High NOTCH3 mRNA levels were found in sion was analyzed by quantitative RT-PCR with two sets of primers MOLT3 cells, while barely detectable NOTCH3 expression characterized encompassing 1–2and3–4 exon boundaries, respectively. SKW3 cells. Notably, low levels of NOTCH3 transcript were detected in CTCF transcript levels were relatively similar in both normal all normal blood cells. blood cells and T-ALL cell lines, irrespectively of NOTCH3 expression The dependence of NOTCH3 expression on DNA methylation of its (Fig. 3A). In contrast BORIS expression appeared to be restricted to T- promoter, colocalizing with a CpG island (Fig. 1B), was investigated by ALL cell lines. However, the presence of BORIS transcript did not corre- bisulfite sequencing in CD4+ T- lymphocytes and in MOLT3 and late with the expression of NOTCH3. Indeed, it appears to be expressed SKW3 T-ALL cells. Fig. 1C shows that NOTCH3 promoter was by both SKW3 cells, which do not express NOTCH3, and MOLT3 cells hypermethylated only in SKW3 cell line characterized by barely detect- (Fig. 3B and C), which express NOTCH3 and show permissive, able levels of mRNA expression, while it was unmethylated in NOTCH3 unmethylated chromatin status of NOTCH3 promoter (Fig. 1C and D). expressing MOLT3 cell line. Intriguingly, the association between pro- To correlate the observed epigenetic chromatin status of NOTCH3 moter methylation and gene expression cannot be extended to normal promoter with the presence of CTCF and/or BORIS, chromatin immuno- blood cells, where NOTCH3 was not expressed in spite of the absence of precipitation (ChIP) experiments were carried out on MOLT3 and SKW3 promoter methylation. T-ALL cells, in comparison with PBMCs. The data above revealed that the silencing of NOTCH3 in normal As shown in Fig. 3D, the ChIP analysis revealed that CTCF binds to the blood cells is independent of DNA methylation, thus suggesting the in- proximal NOTCH3 promoter fragment containing putative CTCF-binding volvement of other mechanism(s). In order to address this issue, sites 1 and 2 in PBMCs and MOLT3 cells, while no CTCF binding was de- our attention was focused on epigenetic changes involving histone tected in SKW3 cells. Concerning BORIS, the protein was associated with methylation. the same region bound by CTCF in MOLT3 cells, which overexpress In fact, it is well known that histone H3 dimethylation and NOTCH3. Conversely, BORIS binding was undetectable in both PBMCs trimethylation at residue K4 (H3K4me2 and H3K4me3) mark transcrip- and SKW3, which do not express NOTCH3. This finally would suggest tional competent chromatin with H3K4me3 particularly enriched in that the binding of BORIS to the NOTCH3 promoter directly correlates actively transcribed promoters, while histone H3 trimethylation at res- with the gene expression. idue K9 (H3K9me3) and trimethylation at residue K27 (H3K27me3) mark constitutive and facultative heterochromatin, respectively. 2.3. NOTCH3 expression correlates with the methylation status of its By chromatin immunoprecipitation (ChIP) analysis, these modifica- promoter and with BORIS expression in normal and cancer cell lines tions were used to investigate the chromatin conformation status of NOTCH3 promoter in cells differentially expressing NOTCH3 (Fig. 1 D). In order to extend the study of the possible direct correlation be- The presence of a high amount of H3K4me2 and H3K4me3 versus tween NOTCH3 expression with the methylation status of its promoter, H3K9me3 and H3K27me3 indicates an active chromatin status in NOTCH3 mRNA level analysis was firstly performed by real-time RT- MOLT3 cells, while high H3K9me3 versus H3K4me2 and H3K4me3 indi- PCR in normal blood cells (PBMCs, CD4+ and CD8+ T-lymphocytes), cates a non-permissive chromatin status in SWK3 cells. These results in immortalized normal human fibroblasts (BJ) and in cell lines derived match with data of RT-PCR and cytosine methylation levels, showing from several cancers, including T-ALL (MOLT3, SKW3 and Jurkat), breast that NOTCH3 is expressed in MOLT3 and silenced in SKW3 cells by con- cancer (MCF7, T47D, MDA-MB-231, MDA-MB-453), melanoma (M14, stitutive heterochromatization. MDA-MB-435), and colon carcinoma (HCT116, HCT15, HT29). Fig. 4A M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 815

Fig. 1. Differential expression and promoter epigenetic status of NOTCH3 in normal blood cells and T-ALL cell lines. (A) Expression of NOTCH3 transcript as detected by quantitative RT-PCR in normal blood cells (CD4+, CD8+ T-lymphocytes and PBMCs,) and in T-ALL cell lines (MOLT3 and SKW3). GUSB transcript level was used as endogenous control. Relative levels were calculated taking the transcriptional levels of NOTCH3 in PBMCs as 1. Results are means ± s.e.m. calculated from three experiments. (B) Schematic map of the NOTCH3 5′-region showing the position of the PCR product analyzed in the assay. TSS indicates the transcriptional start site. Gray box: untranslated 5′-region; open box: coding exon; black horizontal line: CpG island. Numbers in brackets indicate the nucleotide positions relative to the ATG initiator codon. (C) DNA methylation profiling of NOTCH3 promoter as detected by bisulfite sequencing in CD4+ normal blood T-cells and in the leukemic cell lines MOLT3 and SKW3. Each circle represents a CpG dinucleotide. Open circle, unmethylated CpG; black circle, methylated CpG. X indicates cytosine positions where the sequence was indefinable. (D) ChIP analysis of NOTCH3 promoter region for histone H3 modifications in PBMCs and in the MOLT3 and SKW3 T-ALL cell lines. Immunoprecipitation was performed by using specific antibodies to dimethyl and trimethyl lysine 4 (H3K4me2 and H3K4me3) and to trimethyl lysine 27 and 9 (H3K27me3 and H3K9me3). Normal IgG were used as negative control (IgG). Arrow indicates the expected position of the band. Diagram represents the NOTCH3 5′-regionwiththeschematic illustration of the location of the PCR product analyzed in the assay. 816 M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825

Fig. 2. In silico prediction of CTCF-binding motifs in the NOTCH3 promoter region. (A) Schematic representation of the NOTCH3 5′-region with the three sites showing high degree of similarity to CTCF binding sites as predicted by the Genomatix MatInspector software. TSS indicates the transcriptional start site. Gray box: untranslated 5′-region; open box: coding exon. Numbers in brackets indicate the nucleotide positions relative to the ATG initiator codon. (B) Sequence alignment of the putative CTCF sites in the NOTCH3 promoter with the 20-bp core motif of known human CTCF DNA binding sites. Highlighted bold text: NOTCH3 sequence. Matching nucleotides between NOTCH3 and CTCF target genes are highlighted in light gray. Open box indicates nucleotide positions that have been found highly conserved between functional CTCF binding sites genome-wide [20]. shows differential expression of NOTCH3 mRNA levels in different types HEK293T cells. According to BORIS presence, the NOTCH3 promoter activ- of tumors, as well as in immortalized and normal human fibroblasts (BJ) ity was significantly higher in HEK293T cells. To note, this was ascribable and in embryonic kidney HEK293T cells. Normal blood cells are includ- to the proximal promoter which contains putative CTCF binding sites ed as NOTCH3 non-expressing normal cells. (Fig. 5B). The activatory role played by BORIS on NOTCH3 promoter was Then the DNA methylation status of NOTCH3 promoter in different further confirmed by siRNA-mediated BORIS silencing in HEK293T cells cells was investigated by bisulfite PCR followed by restriction analysis (Fig. 5C). Taken together, these results indicate that BORIS, by targeting (COBRA). the proximal region of its promoter, is a positive regulator of NOTCH3. After bisulfite modification and PCR amplification of the NOTCH3 pro- To address in-depth the molecular mechanism by which BORIS is in- moter region, the PCR product was digested with BstUI restriction en- volved in NOTCH3 upregulation, experiments of BORIS silencing/overex- zyme specific to CGCG containing sequences (Fig. 4B). Therefore, pression were performed in the BORIS-positive cell line HEK293T following bisulfite reaction, cleavage occurs selectively when target se- showing unmethylated NOTCH3 promoter and NOTCH3 expression quences are methylated. Fig.4C shows that NOTCH3 promoter was (Fig. 4 and Fig. S3). hypermethylated only in cancer cell lines characterized by its low In agreement with data obtained in luciferase reporter assays mRNA expression (SKW3, HT29, HELA, M14 and MDA-MB-231), while (Fig. 5C) the expression of the endogenous NOTCH3 depended on the it was unmethylated in all NOTCH3-expressing T-ALL and cancer cell BORIS level. In fact, endogenous NOTCH3 was downregulated or upreg- lines and in embryonic HEK293T cells. Notably, normal blood cells and fi- ulated at both mRNA and protein levels following BORIS silencing or broblast cell lines do not express NOTCH3 in spite of the absence of pro- overexpression, respectively (Fig. 6A and B). moter methylation. ChIP analysis experiments allowed us to further study what occurs A parallel analysis was performed on the same cell lines in order to on NOTCH3 promoter following the modification of the BORIS level. assess the presence of BORIS and CTCF transcripts. The results have As shown in Fig. 6C, the silencing of BORIS reduced the amount of pro- shown that while CTCF is expressed at similar levels in all samples tein bound on NOTCH3 promoter, while upregulated BORIS protein levels (Supplementary Fig. S2), BORIS expression is restricted to cancer cell increased BORIS binding. Notably, these binding modifications correlate lines (Fig. 4D and E), irrespective of NOTCH3 expression. with changes in histone methylation patterns: enhanced BORIS occupan- The results at mRNA levels were confirmed at the protein level by cy was associated with an increased amount of H3K4me3, while BORIS Western blots (Supplementary Fig. S3). depletion changed chromatin status of NOTCH3 promoter towards non- permissive chromatin configuration as indicated by the decrease of the 2.4. BORIS positively regulates NOTCH3 transcription and activity H3K4me3 and the introduction of H3K27me3. Overall, these results show a direct relationship between BORIS Together, the results reported above indicate that, while BORIS ex- binding to NOTCH3 promoter, histone methylation status and the expression alone is not sufficient, the combination of unmethylated status pression of NOTCH3. Furthermore, they indicate that the binding of of NOTCH3 promoter and BORIS expression represents the necessary BORIS to NOTCH3 promoter was sufficient to maintain switched on its condition for NOTCH3 expression in cancer cells and in embryonic transcription. HEK293T cells. Consistent with these findings, exogenous BORIS enforced expression Luciferase reporter assay was used to investigate whether and where was sufficient to allow NOTCH3 expression in BORIS-negative BJ normal NOTCH3 promoter activity directly responds to BORIS. Plasmids contain- fibroblasts (Fig. 7A and B). Notably, NOTCH3 upregulation associated ing the NOTCH3 full promoter or its proximal/distal regions (Fig. 5A) with BORIS recruitment to NOTCH3 promoter and with changes of pro- were transfected into both BORIS-negative BJ and BORIS-positive moter chromatin status towards permissive chromatin configuration as M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 817

Fig. 3. CTCF and BORIS differentially occupy NOTCH3 promoter in normal blood cells and T-ALL cell lines. (A) Expression of CTCF transcript as detected by quantitative RT-PCR in normal blood cells (CD4+, CD8+ T-lymphocytes and PBMCs) and in T-ALL cell lines (MOLT3 and SKW3). GUSB transcript level was used as endogenous control. Relative levels were calculated taking the transcriptional levels of CTCF in PBMCs as 1. (B and C) Expression of BORIS transcripts as detected by quantitative RT-PCR in normal blood cells (CD4+, CD8+ T-lymphocytes and PBMCs,) and in T-ALL cell lines (MOLT3 and SKW3). Two sets of primers encompassing 1–2(E1–E2) and 3–4(E3–E4) exons were used to detect all known isoforms of BORIS transcript. GUSB transcript served as loading control of total RNA. Relative levels were calculated taking the transcriptional levels of BORIS in MOLT3 as 1. Results shown in A, B and C are means ± s.e.m. calculated from three experiments. ND, not detectable. (D) In vivo binding of CTCF and BORIS on the 5′-region of NOTCH3 as detected by ChIP assay in PBMCs and in the leukemic cell lines MOLT3 and SKW3. Normal IgG were used as negative ChIP control (IgG). W indicated the no-template PCR control. Diagram represents the NOTCH3 5′-region with the location of the putative CTCF binding sites and the PCR products analyzed in the assay. Region b corresponds to the PCR amplicon indicated in Fig. 1D. TSS indicates the transcriptional start site. Gray box: untranslated 5′-region; open box: coding exon. Numbers in brackets indicate the nucleotide positions relative to the ATG initiator codon. indicated by the increase in H3K4me3 and the depletion in H3K27me3 COBRA assay was carried out in parallel in the same samples, to ver- histone modifications (Fig. 7C). ify the methylation status of NOTCH3 promoter. As shown in Fig. 8Aand To determine the functional relevance of the BORIS-dependent B, sustained levels of NOTCH3 mRNA expression were associated with NOTCH3 upregulation we next investigated whether this event could the unmethylated status of promoter in all analyzed T-ALL samples. initiate the NOTCH3 signaling in BJ cells. To this end, BORIS overexpres- Notably, the samples also showed the presence of the BORIS transcript sion was combined with the treatment with a specific NOTCH3- (Fig. 8C and D), further supporting the hypothesis that aberrant expres- neutralizing antibody able to inhibit the receptor activity. To assess sion of BORIS combined with the unmethylated status of NOTCH3 pro- the NOTCH3 signaling activity, the levels of NOTCH3-IC and of its target moter is the event responsible for NOTCH3 expression in cancer. gene HES1 [1,2] were monitored. We observed that upon BORIS overexpression, the NOTCH3-IC accumulation associated with a significant upregulation of HES1 mRNA. The latter effect was clearly dependent on 3. Discussion NOTCH3 activity as indicated by the fact that it was almost completely abrogated by the addition of the NOTCH3 neutralizing antibody NOTCH3 upregulation is involved in the pathogenesis of and charac- (Fig. 7E), which is also able to reduce the levels of NOTCH3-IC terizes T-ALL [6], although the causes of such an upregulation are poorly (Fig. 7D). These results indicate that the NOTCH3 protein induced by known. Thus, the molecular mechanisms underlying NOTCH3 transcrip- BORIS is functional. tional deregulation in T-ALL cells have been here investigated focusing on epigenetic events and transcriptional factors involved. Our data firstly show that NOTCH3 gene expression does not directly 2.5. Co-expression of BORIS and NOTCH3 in T-ALL correlate with the methylation of its promoter. In fact, while promoter methylated status correlates with NOTCH3 expression in T-ALL cell Considering the association between BORIS expression and NOTCH3 lines, being unmethylated in NOTCH3-expressing MOLT3 cells and upregulation observed in T-ALL cell lines, whether or not this observa- hypermethylated in NOTCH3-negative SKW3 cells, its unmethylated tion could be extended to human primary T-ALL samples was verified. status was not found to be enough to allow NOTCH3 transcriptioninnor- The mRNA levels of NOTCH3 and BORIS were examined in eight cases mal blood cells. This finding indicates that the absence of cytosine meth- of T-ALL and compared with normal peripheral blood T lymphocytes ylation is not the key factor involved in transcriptional upregulation of and with MOLT3 cells. NOTCH3 in T-ALL cell lines and human primary samples, thus suggesting 818 .Zmir ta./Bohmc tBohsc ca13 21)813 (2014) 1839 Acta Biophysica et Biochimica / al. et Zampieri M. – 825

Fig. 4. Analysis of NOTCH3 expression and promoter methylation status and of BORIS expression in normal and cancer cell lines. (A) Expression of NOTCH3 transcript as detected by quantitative RT-PCR. Relative levels were calculated taking the transcriptional levels of NOTCH3 inPBMCsas1.GUSB transcript level was used as endogenous control. Histogram columns are labeled by the mean expression values calculated from three experiments. Error bars indicate ± s.e.m. (B) Schematic map of the NOTCH3 5′-region showing the position of the PCR product analyzed in the COBRA assay. Arrows indicate the approximate position of BstUI sites. TSS indicates the transcriptional start site. Gray box: untranslated 5′-region; open box: coding exon. Numbers in brackets indicate the nucleotide positions relative to the ATG initiator codon. (C) NOTCH3 promoter DNA methylation profiling as detected by COBRA assay. A 332-bp PCR fragment is cleaved by BstUI when methylation is present in bisulfite-modified DNA. UNCUT and CUT indicate samples treated or not with the restriction enzyme. (D and E) Expression of BORIS transcript as detected by quantitative RT-PCR. Two sets of primers encompassing 1–2(E1–E2) and 3–4(E3–E4) exons were used to detect all known isoforms of BORIS transcript. Detection of GUSB transcript served as loading control of total RNA. Relative levels were calculated taking the transcriptional levels of BORIS in MOLT3 as 1. Histogram columns are labeled by the mean expression values calculated from three experiments. Error bars indicate ± s.e.m. M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 819

Fig. 5. BORIS stimulates the activity of proximal NOTCH3 promoter. (A) Schematic map of the NOTCH3 5′-region showing the regions analyzed in the luciferase reporter assays. Numbers in brackets indicate the nucleotide position relative to the ATG initiator codon. (B) Luciferase reporter activity in BORIS-negative BJ and BORIS-positive HEK293T cells. Cells were transfected with luciferase reporter plasmids containing the fragments of NOTCH3 5′-region indicated in A. Cells were also transfected with CMV-β-gal to control transfection efficiency. Results are presented as ratio of luciferase/β-galactosidase relative units and represent the relative fold change compared with the activity of the full promoter construct measured in the BJ sample. (C) Luciferase reporter activity after modulation of BORIS expression levels by siRNA silencing in HEK293T cells. Transfection of non-targeting siRNAs (siRNA CT) was performed as control. Cells were also transfected with CMV-β-gal to control transfection efficiency. Results are presented as ratio of luciferase/β-galactosidase relative units and represent the relative fold change compared with controls. Results shown in B and C are means ± s.e.m. calculated from three experiments. *P b 0.05 compared with BJ (B) or controls (C) (Student's t test). that mechanisms other than DNA methylation are involved in the up- In BORIS-negative normal blood cells, the presence of repressive H3 regulation of NOTCH3 in T-ALL. modifications coincides with CTCF binding at proximal NOTCH3 pro- Additional details were provided by the analysis of histone H3 meth- moter. Conversely, in MOLT3 cells the binding of both CTCF and BORIS ylation, which is known to acts in a combinatorial fashion to specify dis- correlates with an active chromatin status and upregulated NOTCH3 ex- tinct transcriptional chromatin conformation [28]. The DNA promoter pression. To note, in SKW3 cells, characterized by NOTCH3 silencing, hypermethylation of NOTCH3-negative cancer cells (SKW3) associates neither CTCF nor BORIS are bound on the hypermethylated/heterochro- with the predominant presence of H3K9me3, which generally marks si- matic promoter. These results show that NOTCH3 expression varies ac- lenced heterochromatin. Conversely, unmethylated promoter in cording to the binding of BORIS rather than CTCF, and that this binding NOTCH3-positive T-ALL cells (MOLT3) and in normal blood cells depends, in turn, on promoter chromatin accessibility. This suggests shows divergent histone H3 methylation profiles. In NOTCH3-positive that both the unmethylated status of NOTCH3 promoter and BORIS ex- leukemia cells the higher presence of H3K4me3 versus H3K27me3 pression and binding to the promoter are required for NOTCH3upregu- clearly indicate transcriptional permissive status of the promoter, lation in T-ALL and cancer cell lines. Consistent with this hypothesis, we while the higher level of H3K27me3 versus H3K4me3 in normal blood found that NOTCH3 upregulation occurs in BORIS-positive cell lines de- cells explains the silencing of NOTCH3 despite the unmethylated status rived from different cancers, showing unmethylated NOTCH3 promoter. of promoter. Thus, different from DNA methylation, H3 histone methyl- To obtain information about the direct role played by BORIS in ation seems to discriminate between normal and leukemia cells. NOTCH3 expression the highly transfectable NOTCH3- and BORIS-ex- The involvement of epigenetic factors in driving chromatin confor- pressing HEK293T cell line was used. Promoter activity assays in the mation changes in the NOTCH3 locus was then investigated. Attention condition of silencing BORIS indicate that it is directly involved in the was focused on CTCF [29–32] and BORIS/CTCFL [16,17] paralog proteins, transcriptional upregulation of NOTCH3. Moreover, a direct analysis of as binding sites for CTCF were identified on NOTCH3 promoter. BORIS is what occurred on endogenous NOTCH3 locus following modulation of very similar to CTCF in the 11-zinc-finger DNA-binding domain and, the BORIS level allowed us to establish that BORIS is directly responsible sharing part of DNA target sequences with this protein, could compete for changes in the chromatin structure necessary for transcriptional ac- with it. Conversely, CTCF and BORIS diverge in the N-and C-terminal tivation of NOTCH3. In fact, the presence of BORIS on NOTCH3 promoter ends, allowing interaction with distinct partners [17,24–26]. associates with an increased H3K4me3/H3K27me3 ratio. BORIS is a cancer testis antigen (CTA) as it is normally expressed in Confirmatory finding of a role for BORIS in NOTCH3 expression was testis but aberrantly expressed in several primary tumors and established obtained in the BORIS- and NOTCH3-negative BJ normal fibroblast cancer cell lines [16,17,27,33–35]. Importantly, BORIS itself has a pivotal where exogenous BORIS enforced expression is sufficient to induce role in the expression of other CTAs (e.g. MAGE-A1) [36–39] and of NOTCH3 upregulation. some proto-oncogenes (e.g. hTERT) [40,41], generally repressed by CTCF. Our results are in agreement with literature data showing that BORIS Analysis of CTCF and BORIS expression shows that, while CTCF is positively regulates gene expression by inducing an open chromatin expressed in both normal blood cells and T-ALL cell lines, BORIS is spe- conformation [41–44]. BORIS induced alterations in the NOTCH3 pro- cifically expressed in T-ALL and solid tumor-derived cell lines. moter appear highly similar to those common features typical of permis- Moreover, ChIP assay, used to assess the binding of CTCF and BORIS sive/active chromatin conformation, which have recently been reported proteins to the 5′ region of the NOTCH3 gene, demonstrated that the oc- for Gal3st1 [43], in the physiological context of spermatogenesis, and for cupancy of NOTCH3 promoter by CTCF and/or BORIS correlates with the activation of cancer testis antigen TSP50 [39]. To note, genome-wide NOTCH3 expression, promoter epigenetic status and BORIS expression. analyses revealed that BORIS co-localizes with H3K4me3 and active PolII 820 M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825

Fig. 6. BORIS is a positive epigenetic regulator of NOTCH3 expression. Transient transfection of either siRNA (siRNA BORIS) or expression vectors for BORIS (Myc-BORIS) was performed in HEK293T cells. Transfection of non-targeting siRNAs (siRNA CT) and empty overexpression vector (Vector) was performed as control. (A) Western blot analysis of cell lysates with anti- BORIS, anti-CTCF and anti-NOTCH3 antibodies were performed to monitor endogenous protein levels. Anti-Myc antibodies were used to detect exogenous BORIS protein. Anti α-tubulin antibodies served as control for loading differences. (B) Expression of NOTCH3 transcript as detected by quantitative RT-PCR in HEK293T cells after transfection of either siRNA or expression vectors for BORIS. Relative levels were calculated taking the transcriptional levels of NOTCH3 in control transfection as 1. GUSB transcript level was used as endogenous control. Results are means ± s.e.m. calculated from three experiments. (C) ChIP analysis of NOTCH3 promoter region for Histone H3 modifications in HEK293T cells after transfection of either siRNA or expression vectors for BORIS. Immunoprecipitation was performed by using specific antibodies to trimethyl lysine 4 (H3K4me3) and to trimethyl lysine 27 (H3K27me3). Normal IgG were used as negative control (IgG). The DNA region analyzed corresponds to the PCR amplicon indicated in Fig. 1D. *P b 0.05 compared with controls (Student's t test). on the TSS of active promoters [26]. This is in agreement with data show- Furthermore, this study supports the opportunity to exploit BORIS ing that SET1A H3K4 methyltransferase is a partner of BORIS [45]. targeting for anti-cancer therapy [47–50]. Considering that abnormal BORIS expression has been correlated with the upregulation of proto-oncogenes in different primary tumors 5. Materials and methods [35], our next aim was to gain evidence that, in T-ALL primary samples, where enhanced NOTCH3 gene expression was found [6],BORISwasco- 5.1. Cell culture, isolation of peripheral blood cells and patient samples expressed. We found that BORIS expression correlated with the upregulation of NOTCH3 in all inspected T-ALL patient samples. Experiments were carried out in human normal foreskin fibroblasts expressing hTERT (BJ); T-cell acute lymphoblastic leukemia cell lines 4. Conclusions MOLT3, SKW3 and Jurkat; breast cancer cell lines MCF7, T47D, MDA- MB-231 and MDA-MB-453; melanoma cell lines M14 and MDA-MB- Taken together our data suggest that BORIS aberrant expression in T- 435; colorectal cancer cell line HCT116, HCT15 and HT29; cervical can- ALL triggers NOTCH3 transcriptional activation most likely through an cer cell line HeLa; glioblastoma cell line U87; embryonic kidney cell epigenetic mechanism. By binding unmethylated CTCF binding sites line HEK293T. Cell lines were maintained as subconfluent culture in present on NOTCH3 promoter, BORIS favors an active chromatin confor- high-glucose DMEM (adherent cell culture) or RPMI (suspension cell mation by increasing the H3K4me3/H3K27me3 ratio (Fig. 9). Anyway, culture) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, even though our data suggest that in cancer cell lines the NOTCH3 50 units/ml penicillin and 50 mg/ml streptomycin. All culture solutions promoter-region is co-bound simultaneously by both CTCF and BORIS, were from VWR International PBI (Milan, Italy). other alternative patterns of DNA-occupancy could not be ruled out. Peripheral blood mononuclear cells (PBMCs) were obtained from In fact, this region contains at least two putative target sites for the blood of normal donors by density gradient centrifugation using 11-zinc-finger DNA-binding domains to be independently bound either ‘Lymphoprep’ solution (Axis-Shield, Dundee, UK). by CTCF or by BORIS. For lymphocytes preparation cells were obtained as previously de- In conclusion, this paper not only provides insights into molecular scribed [51]. Briefly, the mature T cells, including CD8 single-positive mechanisms underlying NOTCH3 upregulation in T-ALL, but also and CD4 single-positive, were positively selected by magnetic cell sepa- suggests a novel molecular model for the action of BORIS that could be ration by using the MACS (magnetic cell sorter) system (Milteny Biotec, crucial in explaining the activation of CTAs and other proto-oncogenes Auburn, CA) in accordance with the manufacturer's instructions. The in T-ALL [46]. purity of the isolated fractions was greater than 90%, as determined by M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 821

Fig. 7. BORIS enforced expression is sufficient to induce NOTCH3 upregulation in BORIS-negative BJ normal fibroblasts. (A) Western blot analysis of cell lysates after transfection of BORIS expression vector (Myc-BORIS) and empty overexpression vector (Vector) as control. HEK293T samples served as positive control for BORIS and NOTCH3. Analysis was performed with anti-BORIS, anti-NOTCH3 and anti-CTCF antibodies. Anti-Myc antibodies were used to detect exogenous BORIS protein. Anti α-tubulin antibodies served as control for loading differences. (B) Expression of NOTCH3 transcript as detected by quantitative RT-PCR after transfection of BORIS expression vector. Relative levels were calculated taking the transcriptional levels of NOTCH3 in control transfection as 1. GUSB transcript level was used as endogenous control. Results are means ± s.e.m. calculated from three experiments. (C) ChIP analysis of NOTCH3 promoter region for Histone H3 modifications after transfection of expression vectors for BORIS. Immunoprecipitation was performed by using specific antibodies to trimethyl lysine 4 (H3K4me3) and to trimethyl lysine 27 (H3K27me3). Normal IgG were used as negative control (IgG). The DNA region analyzed corresponds to the PCR amplicon indicated in Fig. 1D. (D) Western blot analysis of cell lysates after transfection of BORIS expression vector and treatment with an anti-NOTCH3 IgG, known to neutralize receptor function (Ab-N3), or with the normal control IgG (IgG). Analysis was performed with anti-Myc antibodies to detect exogenous BORIS protein and with anti-NOTCH3 antibodies. Anti α-tubulin antibodies served as control for loading differences. (E) Expression of HES-1 transcript as detected by quantitative RT-PCR after transfection of BORIS expression vector and treatment with the anti- NOTCH3 neutralizing IgG. Relative levels were calculated taking the transcriptional levels of HES-1 in control transfections as 1. GUSB transcript level was used as endogenous control. Re- sults are means ± s.e.m. calculated from three experiments. *P b 0.05 compared with controls (Student's t test).

ﬂow cytometry analysis (Becton Dickinson, Franklin Lakes, NJ). Anti- Germantown, MD). Concentration, purity and integrity of preparations CD4-FITC, anti-CD8-PE, anti-CD122-FITC, and anti-CD45R/B220-PE Abs were evaluated by spectrophotometry, followed by agarose gel– were purchased from Pharmingen (San Diego, CA). ethidium bromide electrophoresis. Eight primary T-ALL samples were obtained at the onset of disease. All samples contained at least 70% leukemic cells. 5.3. Reverse transcription-PCR (RT-PCR) 5.2. Extraction of nucleic acids Total RNA (1 μg) was subjected to reverse transcription using the Genomic DNA was prepared by the DNeasy tissue kit (Qiagen, Ger- SuperScript VILO cDNA synthesis kit (Life Technologies, Invitrogen, mantown, MD). Total RNA was puriﬁed by RNeasy mini kit (Qiagen, Carlsbad, CA). 822 M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825

Fig. 8. Co-expression of BORIS and NOTCH3 in T-ALL primary samples. (A) Expression of NOTCH3 transcript as detected by quantitative RT-PCR in normal blood CD4+ and CD8+ T-cells, in eight cases of T-ALL, and in the leukemic cell lines MOLT3. Relative levels were calculated taking the transcriptional level of NOTCH3 in CD4+ T-cells as 1. GUSB transcript level was used as endogenous control. Histogram columns are labeled by the mean expression values calculated from three experiments. Error bars indicate ± s.e.m. (B) NOTCH3 promoter DNA methylation proﬁling as detected by COBRA assay. CD4+ T-cells and SKW3 served as positive and negative controls, respectively. The DNA region analyzed is indicated in Fig. 4B. (C and D) Ex- pression of BORIS transcript as detected by quantitative RT-PCR in normal blood CD4+ and CD8+ T-cells, in eight cases of T-ALL, and in the leukemic cell lines MOLT3. Two sets of primers encompassing 1–2(E1–E2) and 3–4(E3–E4) exons were used to detect all known isoforms of BORIS transcript. Detection of GUSB transcript served as endogenous control. Relative levels were calculated taking the transcriptional levels of BORIS in MOLT3 as 1. Histogram columns are labeled by the mean expression values calculated from three experiments. Error bars indicate ± s.e.m.

Expression of mRNAs for NOTCH3, CTCF and BORIS was measured by plates on the iCycler IQ detection system (Bio-Rad, Hercules, CA). Nor- real time PCR using Taq-Man gene expression assays (Life Technologies, malization was carried out using beta-glucuronidase (GUSB) as internal Applied Biosystems, Carlsbad, CA) following the manufacturer's protocol control gene. TaqMan gene expression assay IDs for each set of primers for the comparative CT method. The ampliﬁcation reaction was per- and probe were: Hs01128541_m1 (NOTCH3); Hs00902008_m1 (CTCF); formed using the TaqMan Universal PCR master Mix (Life Technologies, Hs00966555_g1 (BORIS E1–E2); Hs00966548_g1 (BORIS E3–E4), Invitrogen, Carlsbad, CA). All samples were run in duplicate in 96-well Hs00172878_m1(HES1) and Hs99999908_m1 (GUSB).

Fig. 9. Aberrant expression of BORIS in T-ALL cells favors NOTCH3 upregulation. Figure shows a possible molecular mechanism underlying the association between BORIS expression and NOTCH3 upregulation in T-ALL cells. The aberrant expression of BORIS in T-ALL cells and its recruitment on unmethylated CTCF binding sites is permissive for the transcription of NOTCH3. In fact, BORIS is required to maintain a local transcriptionally active chromatin conﬁguration at the NOTCH3 promoter. M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 823

5.4. Genomic bisulfite sequencing and combined bisulfite restriction cloned into TA-Cloning cloning vector (Life Technologies, Invitrogen, analysis (COBRA) Carlsbad, CA) and a KpnI and XhoI fragment containing the promoter was isolated and subcloned into polylinker of the plasmid PGL3 Basic Bisulfite treatment of genomic DNA (1 μg) was performed by using vector (Promega, Madison, WI). This construct was denoted as full the EZ DNA methylation kit (Zymo Research, Irvine, CA). length NOTCH3-promoter. Deletion fragments of the NOTCH3 promoter For sequencing analyses the NOTCH3 promoter region was obtained were generated using HindIII restriction site present within the by PCR. The amplification reactions were carried out by using the NOTCH3-promoter which resulted in a distal fragment (−1274 to AccuPrime Taq DNA polymerase system (Life Technologies, Invitrogen, −403 bp) and a proximal fragment (−404 to +259 bp) containing Carlsbad, CA) with 10 pmol specific primers and 100 ng of bisulfite- CTCF sites. treated DNA in a finalreactionmixof25μl. The following cycling condi- Boris cDNA, containing the complete coding sequence (accession no. tions were adopted. The initial denaturation at 94 °C for 2 min; 40 cy- DQ153171), was isolated by PCR amplification using the template cDNA cles of: 94 °C for 30 s, 52 °C for 30 s, and 72 °C for 30 s. Sequencing prepared from mouse adult testis. The following oligonucleotides were analysis was performed as described previously [52]. used: 5′-CACGGAATTCAATGGCAGCCACTGAGATCTCTGTCCTT-3′sense, For COBRA analyses the NOTCH3 promoter region was obtained by 5′-CCTGCTCGAGTCACTTATCCATCGTGTTGAGGAGCAT-3′ antisense. The nested PCR. PCR amplified fragment was cloned in the myc-tag expression vector The amplification reactions were carried out by using the AccuPrime pCS2-MT that inserts 6 myc tags in-frame at the carboxyl terminus of Taq DNA polymerase system (Life Technologies, Invitrogen, Carlsbad, the protein and sequenced by Eurofins MWG Operon service. CA) with 10 pmol specific primers and 100 ng of bisulfite-treated DNA For the luciferase assays cells were transfected with Lipofectamine (first amplification) or 2 μlofthefirst amplification in a final reaction 2000 (Life Technologies, Invitrogen, Carlsbad, CA) according to the mix of 25 μl. manufacturer's instructions. Cells were seeded in 12-well culture plates The following cycling conditions were adopted. The initial denatur- at a concentration of 2 × 105 cells/well and incubated for 24 h prior to ation at 94 °C for 2 min; 45 cycles of: 94 °C for 30 s, 50 °C (first ampli- cotransfections involving luciferase reporter constructs (30 ng/well), fication) or 52 °C (second amplification) for 30 s, and 72 °C for 40 s. anti-BORIS (siGENOME SMARTpool, Thermo Fisher Scientific, 30 pmol/ Enzyme digestion was performed with BstUI restriction endonuclease well) or control siRNAs (siGENOME Non-Targeting siRNA Pool, Thermo (New England Biolabs, Ipswich, MA) according to manufacturer's proto- Fisher Scientific, 30 pmol/well). As internal control, plasmid pCMV- col on 5 μl of the second amplification product in a final reaction mix of beta-gal (0.5 ng/well) (Promega, Madison, WI), containing the beta- 20 μl. Agarose gel electrophoresis (2%) followed by ethidium bromide galactosidase reporter gene, was co-transfected. Luciferase and beta- staining was performed to visualize digestion products (20 μl). Primers galactosidase activities were quantitated 24 h later by the luciferase or used for the amplifications are listed in the Supplementary Table. the β-galactosidase enzyme assay system (Promega, Madison, WI), respectively. 5.5. Chromatin immunoprecipitation (ChIP) For the ChIP assays 7 × 105 cells were seeded in 100 × 15 mm culture dishes and incubated for 24 h prior to transfection. Transfection of ChIP was performed as described previously [52,53]. The cross-linking siRNAs targeting BORIS and of control siRNAs (600 pmol/dish) was procedure was adapted for cells growing in suspension as follows. performed with Lipofectamine 2000 reagent using the manufacturer's Starting sample consisted of 1 × 107 cells from subconfluent cultures. protocol. Transfection of myc-Boris construct or empty myc-vector After centrifugation cells were resuspended in complete culture medium (8 μg/dish) was performed with Lipofectamine Plus reagent (Life Tech- added of formaldehyde (Sigma-Aldrich, St. Louis, MO) to a final concen- nologies, Invitrogen, Carlsbad, CA) using the manufacturer's protocol. tration of 1% and incubated 15 min at room temperature with gentle After 48 h the cells were processed for ChIP, RT-PCR and for Western shaking. To stop cross-linking reaction, cells were recovered by centrifu- blot analyses. gation, resuspended in 0.125 M glycine (Sigma-Aldrich, St. Louis, MO) In the NOTCH3 activity downregulation assays, cells were seeded at in PBS and incubated for 5 min at room temperature with gentle shaking. 1.3 × 105 cells per well in 24-well plates 24 h prior to transfections in- The antibodies used for immunoprecipitations were: rabbit poly- volving the myc-Boris construct or empty myc-vector (0.2 μg/well). clonal anti-trimethyl-Histone H3 (Lys9) (Millipore, Billerica, MA); rab- The transfection by Lipofectamine Plus reagent followed the manufac- bit polyclonal anti-trimethyl-Histone H3 (Lys27) (Millipore, Billerica, turers' protocol. At the conclusion of the transfection procedure, cells MA); mouse monoclonal anti-trimethyl-Histone H3 (Lys4) (Millipore, were shifted to growth medium containing 10 μg/ml of either a sheep Billerica, MA); rabbit polyclonal anti-dimethyl-Histone H3 (Lys4) polyclonal anti-NOTCH3 IgG, known to neutralize receptor function (Millipore, Billerica, MA); rabbit polyclonal anti-CTCF (Millipore, Biller- (R&D Systems, Minneapolis, MN), or an equivalent concentration of ica,MA); goat polyclonal anti-BORIS (C-15) (Santa Cruz Biotechnology, the normal IgG sheep control (R&D Systems, Minneapolis, MN) and cul- Santa Cruz, CA) and normal rabbit/mouse IgGs (Santa Cruz Biotechnol- tured for 48 h. Samples were than processed for Western blot and RT- ogy, Santa Cruz, CA). PCR analyses. End-point PCR reactions were carried out by using the AccuPrime GC-Rich DNA polymerase system (Life Technologies, Invitrogen, Carls- 5.7. Western blot bad, CA), with 10 pmol specific primers and 5 μl of immunoprecipitated DNA in a final reaction mix of 25 μl. The following cycling conditions Western blot assays were performed as described elsewhere [54]. were adopted. The initial denaturation at 94 °C for 2 min; 30–45 cycles The antibodies used were the following: mouse monoclonal anti-Myc of: 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s. Agarose gel electro- (9E10 clone, hybridoma-conditioned medium), mouse monoclonal phoresis (1.8%) followed by ethidium bromide staining was performed anti-BORIS (Millipore, Billerica, MA), rabbit polyclonal anti-CTCF to visualize PCR products (8 μl). Primers used for the amplifications are (Millipore, Billerica, MA); rabbit polyclonal anti-NOTCH3 (Cell Signaling listed in the Supplementary table. Technology, Danvers, MA) and mouse monoclonal anti-Tubulin-α (Sigma-Aldrich, St. Louis, MO). 5.6. Plasmid construction and transfection of cells 5.8. Statistical analysis Human NOTCH3 promoter-luciferase reporter plasmids were generated using the following primers: forward, 5-ATCATGCCACTGCACTTC Statistical tests and the number of replicates performed are reported AG-3, and reverse, 5-GTTCTTGCACTCCCCCTCTG-3; a 1533 bp fragment in figure legends. Data were considered to be statistically significant if was amplified from human genomic DNA; the amplified insert was ⁎P b 0.05. 824 M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825

Supplementary data to this article can be found online at http://dx. [21] L.R. Meyer, A.S. Zweig, A.S. Hinrichs, D. Karolchik, R.M. Kuhn, M. Wong, et al., The UCSC Genome Browser database: extensions and updates 2013, Nucleic Acids Res. doi.org/10.1016/j.bbagrm.2014.06.017. 41 (2012) D64–D69. [22] W.J. Kent, C.W. Sugnet, T.S. Furey, K.M. Roskin, T.H. Pringle, A.M. Zahler, et al., The Funding human genome browser at UCSC, Genome Res. 12 (2002) 996–1006. [23] K.R. Rosenbloom, C.A. Sloan, V.S. Malladi, T.R. Dreszer, K. Learned, V.M. Kirkup, et al., ENCODE data in the UCSC Genome Browser: year 5 update, Nucleic Acids Res. 41 This research was supported by grants from the Italian Ministry of (2012) D56–D63. University and Research (MIUR) (P.C.: FIRB-RBIN06E9Z8_003, I.S.: [24] R. de Necochea-Campion, A. Ghochikyan, S.F. Josephs, S. Zacharias, E. Woods, F. FIRB-RBAP11WCRZ, I.S.: PRIN-2010MCLPLB), Sapienza University of Karimi-Busheri, et al., Expression of the epigenetic factor BORIS (CTCFL) in the human genome, J. Transl. Med. 9 (2011) 213. Rome (M.Z.: C26A134WJ2, M.Z.: C26A12PN9T) and the Italian Associa- [25] T.A. Hore, J.E. Deakin, J.A. Marshall Graves, The evolution of epigenetic regulators tion for Cancer Research (AIRC) (I.S.:IG 13314). CTCF and BORIS/CTCFL in amniotes, PLoS Genet. 4 (2008) e1000169. [26] F. Sleutels, W. Soochit, M. Bartkuhn, H. Heath, S. Dienstbach, P. Bergmaier, et al., The male germ cell gene regulator CTCFL is functionally different from CTCF and binds Acknowledgements CTCF-like consensus sites in a nucleosome composition-dependent manner, Epige- netics Chromatin 5 (2012) 8. The authors acknowledge the assistance of Dr Fulvia Brugnoletti [27] E.M. Pugacheva, T. Suzuki, S.D. Pack, N. Kosaka-Suzuki, J. Yoon, A.A. Vostrov, et al., The structural complexity of the human BORIS gene in gametogenesis and cancer, (Department of Cellular Biotechnologies and Hematology, Sapienza PLoS ONE 5 (2010) e13872. University of Rome) in carrying out the isolation of primary T-ALL [28] Z. Wang, C. Zang, J.A. Rosenfeld, D.E. Schones, A. Barski, S. Cuddapah, et al., Combi- samples. natorial patterns of histone acetylations and methylations in the human genome, Nat. Genet. 40 (2008) 897–903. [29] R. Ohlsson, M. Bartkuhn, R. Renkawitz, CTCF shapes chromatin by multiple mecha- References nisms: the impact of 20 years of CTCF research on understanding the workings of chromatin, Chromosoma 119 (2010) 351–360. [1] E.R. Andersson, R. Sandberg, U. Lendahl, Notch signaling: simplicity in design, versa- [30] R. Ohlsson, V. Lobanenkov, E. Klenova, Does CTCF mediate between nuclear organi- tility in function, Development 138 (2011) 3593–3612. zation and gene expression? Bioessays 32 (2010) 37–50. [2] S.J. Bray, Notch signalling: a simple pathway becomes complex, Nat. Rev. 7 (2006) [31] J.E. Phillips, V.G. Corces, CTCF: master weaver of the genome, Cell 137 (2009) 678–689. 1194–1211. [3] K.G. Leong, A. Karsan, Recent insights into the role of Notch signaling in tumorigen- [32] J. Zlatanova, P. Caiafa, CCCTC-binding factor: to loop or to bridge, Cell. Mol. Life Sci. esis, Blood 107 (2006) 2223–2233. 66 (2009) 1647–1660. [4] M. Roy, W.S. Pear, J.C. Aster, The multifaceted role of Notch in cancer, Curr. Opin. [33] V. D'Arcy, Z.K. Abdullaev, N. Pore, F. Docquier, V. Torrano, I. Chernukhin, et al., The Genet. Dev. 17 (2007) 52–59. potential of BORIS detected in the leukocytes of breast cancer patients as an early [5] D. Bellavia, A.F. Campese, E. Alesse, A. Vacca, M.P. Felli, A. Balestri, et al., Constitutive marker of tumorigenesis, Clin. Cancer Res. 12 (2006) 5978–5986. activation of NF-kappaB and T-cell leukemia/lymphoma in Notch3 transgenic mice, [34] V. D'Arcy, N. Pore, F. Docquier, Z.K. Abdullaev, I. Chernukhin, G.X. Kita, et al., BORIS, a EMBO J. 19 (2000) 3337–3348. paralogue of the transcription factor, CTCF, is aberrantly expressed in breast tu- [6] D. Bellavia, A.F. Campese, S. Checquolo, A. Balestri, A. Biondi, G. Cazzaniga, et al., mours, Br. J. Cancer 98 (2008) 571–579. Combined expression of pTalpha and Notch3 in T cell leukemia identifies the re- [35] I. Martin-Kleiner, BORIS in human cancers — a review, Eur. J. Cancer 48 (2012) quirement of preTCR for leukemogenesis, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 929–935. 3788–3793. [36] J.A. Hong, Y. Kang, Z. Abdullaev, P.T. Flanagan, S.D. Pack, M.R. Fischette, et al., Recip- [7] P. Sansone, G. Storci, C. Giovannini, S. Pandolfi, S. Pianetti, M. Taffurelli, et al., rocal binding of CTCF and BORIS to the NY-ESO-1 promoter coincides with derepres- p66Shc/Notch-3 interplay controls self-renewal and hypoxia survival in human sion of this cancer-testis gene in lung cancer cells, Cancer Res. 65 (2005) stem/progenitor cells of the mammary gland expanded in vitro as mammospheres, 7763–7774. Stem Cells 25 (2007) 807–815. [37] S. Renaud, E.M. Pugacheva, M.D. Delgado, R. Braunschweig, Z. Abdullaev, D. [8] S.G. Jung, Y.D. Kwon, J.A. Song, M.J. Back, S.Y. Lee, C. Lee, et al., Prognostic significance Loukinov, et al., Expression of the CTCF-paralogous cancer-testis gene, brother of of NOTCH3 gene expression in ovarian serous carcinoma, Cancer Sci. 101 (2010) the regulator of imprinted sites (BORIS), is regulated by three alternative promoters 1977–1983. modulated by CpG methylation and by CTCF and p53 transcription factors, Nucleic [9] J. van Nes, A. Chan, T. van Groningen, P. van Sluis, J. Koster, R. Versteeg, A NOTCH3 Acids Res. 35 (2007) 7372–7388. transcriptional module induces cell motility in neuroblastoma, Clin. Cancer Res. 19 [38] S. Vatolin, Z. Abdullaev, S.D. Pack, P.T. Flanagan, M. Custer, D.I. Loukinov, et al., Con- (2013) 3485–3494. ditional expression of the CTCF-paralogous transcriptional factor BORIS in normal [10] Y.Z. Ye, Z.H. Zhang, X.Y. Fan, X.L. Xu, M.L. Chen, B.W. Chang, et al., NOTCH3 overex- cells results in demethylation and derepression of MAGE-A1 and reactivation of pression associates with poor prognosis in human non-small-cell lung cancer, Med. other cancer-testis genes, Cancer Res. 65 (2005) 7751–7762. Oncol. 30 (2013) 595. [39] N. Kosaka-Suzuki, T. Suzuki, E.M. Pugacheva, A.A. Vostrov, H.C. Morse 3rd, D. [11] M.A. Alqudah, S. Agarwal, M.S. Al-Keilani, Z.A. Sibenaller, T.C. Ryken, M. Assem, Loukinov, et al., Transcription factor BORIS (Brother of the Regulator of Imprinted NOTCH3 is a prognostic factor that promotes glioma cell proliferation, migration Sites) directly induces expression of a cancer-testis antigen, TSP50, through regulat- and invasion via activation of CCND1 and EGFR, PLoS ONE 8 (2013) e77299. ed binding of BORIS to the promoter, J. Biol. Chem. 286 (2011) 27378–27388. [12] T.P. Dang, A.F. Gazdar, A.K. Virmani, T. Sepetavec, K.R. Hande, J.D. Minna, et al., Chro- [40] S. Renaud, D. Loukinov, L. Alberti, A. Vostrov, Y.W. Kwon, F.T. Bosman, et al., BORIS/ mosome 19 translocation, overexpression of NOTCH3, and human lung cancer, J. CTCFL-mediated transcriptional regulation of the hTERT telomerase gene in testicu- Natl. Cancer Inst. 92 (2000) 1355–1357. lar and ovarian tumor cells, Nucleic Acids Res. 39 (2011) 862–873. [13] M. Kadota, M. Sato, B. Duncan, A. Ooshima, H.H. Yang, N. Diaz-Meyer, et al., Identi- [41] L. Sun, L. Huang, P. Nguyen, K.S. Bisht, G. Bar-Sela, A.S. Ho, et al., DNA methyltrans- fication of novel gene amplifications in breast cancer and coexistence of gene ampli- ferase 1 and 3B activate BAG-1 expression via recruitment of CTCFL/BORIS and mod- fication with an activating mutation of PIK3CA, Cancer Res. 69 (2009) 7357–7365. ulation of promoter histone methylation, Cancer Res. 68 (2008) 2726–2735. [14] F. Micci, I. Panagopoulos, G.E. Tjonnfjord, A. Kolstad, J. Delabie, K. Beiske, et al., Mo- [42] S. Bhan, S.S. Negi, C. Shao, C.A. Glazer, A. Chuang, D.A. Gaykalova, et al., BORIS bind- lecular cytogenetic characterization of t(14;19)(q32;p13), a new recurrent translo- ing to the promoters of cancer testis antigens, MAGEA2, MAGEA3, and MAGEA4, is cation in B cell malignancies, Virchows Arch. 450 (2007) 559–565. associated with their transcriptional activation in lung cancer, Clin. Cancer Res. 17 [15] J.T. Park, M. Li, K. Nakayama, T.L. Mao, B. Davidson, Z. Zhang, et al., NOTCH3 gene (2011) 4267–4276. amplification in ovarian cancer, Cancer Res. 66 (2006) 6312–6318. [43] T.Suzuki,N.Kosaka-Suzuki,S.Pack,D.M.Shin,J.Yoon,Z.Abdullaev,etal.,Ex- [16] E.M. Klenova, H.C. Morse 3rd, R. Ohlsson, V.V. Lobanenkov, The novel BORIS + CTCF pression of a testis-specific form of Gal3st1 (CST), a gene essential for spermato- gene family is uniquely involved in the epigenetics of normal biology and cancer, genesis, is regulated by the CTCF paralogous gene BORIS, Mol. Cell. Biol. 30 Semin. Cancer Biol. 12 (2002) 399–414. (2010) 2473–2484. [17] D.I. Loukinov, E. Pugacheva, S. Vatolin, S.D. Pack, H. Moon, I. Chernukhin, et al., [44] D. Gaykalova, R. Vatapalli, C.A. Glazer, S. Bhan, C. Shao, D. Sidransky, et al., Dose- BORIS, a novel male germ-line-specific protein associated with epigenetic dependent activation of putative oncogene SBSN by BORIS, PLoS ONE 7 (2012) reprogramming events, shares the same 11-zinc-finger domain with CTCF, the insu- e40389. lator protein involved in reading imprinting marks in the soma, Proc. Natl. Acad. Sci. [45] P. Nguyen, G. Bar-Sela, L. Sun, K.S. Bisht, H. Cui, E. Kohn, et al., BAT3 and SET1A form U. S. A. 99 (2002) 6806–6811. a complex with CTCFL/BORIS to modulate H3K4 histone dimethylation and gene ex- [18] K. Cartharius, K. Frech, K. Grote, B. Klocke, M. Haltmeier, A. Klingenhoff, et al., pression, Mol. Cell. Biol. 28 (2008) 6720–6729. MatInspector and beyond: promoter analysis based on transcription factor binding [46] H. Nishikawa, Y. Maeda, T. Ishida, S. Gnjatic, E. Sato, F. Mori, et al., Cancer/testis an- sites, Bioinformatics 21 (2005) 2933–2942. tigens are novel targets of immunotherapy for adult T-cell leukemia/lymphoma, [19] K. Quandt, K. Frech, H. Karas, E. Wingender, T. Werner, MatInd and MatInspector: Blood 119 (2012) 3097–3104. new fast and versatile tools for detection of consensus matches in nucleotide se- [47] A. Ghochikyan, M. Mkrtichyan, D. Loukinov, G. Mamikonyan, S.D. Pack, N. quence data, Nucleic Acids Res. 23 (1995) 4878–4884. Movsesyan, et al., Elicitation of T cell responses to histologically unrelated tumors [20] R.N. Plasschaert, S. Vigneau, I. Tempera, R. Gupta, J. Maksimoska, L. Everett, et al., by immunization with the novel cancer-testis antigen, brother of the regulator of CTCF binding site sequence differences are associated with unique regulatory and imprinted sites, J. Immunol. 178 (2007) 566–573. functional trends during embryonic stem cell differentiation, Nucleic Acids Res. 42 [48] D. Loukinov, A. Ghochikyan, M. Mkrtichyan, T.E. Ichim, V.V. Lobanenkov, D.H. Cribbs, (2014) 774–789. et al., Antitumor efficacy of DNA vaccination to the epigenetically acting tumor M. Zampieri et al. / Biochimica et Biophysica Acta 1839 (2014) 813–825 825

promoting transcription factor BORIS and CD80 molecular adjuvant, J. Cell. Biochem. [52] M. Zampieri, T. Guastafierro, R. Calabrese, F. Ciccarone, M.G. Bacalini, A. Reale, et al., 98 (2006) 1037–1043. ADP-ribose polymers localized on Ctcf-Parp1-Dnmt1 complex prevent methylation [49] M. Mkrtichyan, A. Ghochikyan, H. Davtyan, N. Movsesyan, D. Loukinov, V. of Ctcf target sites, Biochem. J. 441 (2012) 645–652. Lobanenkov, et al., Cancer-testis antigen, BORIS based vaccine delivered by dendrit- [53] M. Zampieri, C. Passananti, R. Calabrese, M. Perilli, N. Corbi, F. De Cave, et al., Parp1 ic cells is extremely effective against a very aggressive and highly metastatic mouse localizes within the Dnmt1 promoter and protects its unmethylated state by its en- mammary carcinoma, Cell. Immunol. 270 (2011) 188–197. zymatic activity, PLoS ONE 4 (2009) e4717. [50] M. Mkrtichyan, A. Ghochikyan, D. Loukinov, H. Davtyan, T.E. Ichim, D.H. Cribbs, et al., [54] T. Guastafierro, A. Catizone, R. Calabrese, M. Zampieri, O. Martella, M.G. Bacalini, DNA, but not protein vaccine based on mutated BORIS antigen significantly inhibits et al., ADP-ribose polymer depletion leads to nuclear Ctcf re-localization and chro- tumor growth and prolongs the survival of mice, Gene Ther. 15 (2008) 61–64. matin rearrangement, Biochem. J. 449 (2013) 623–630. [51] C. Talora, S. Cialfi, C. Oliviero, R. Palermo, M. Pascucci, L. Frati, et al., Cross talk among NOTCH3, pre-TCR, and Tal1 in T-cell development and leukemogenesis, Blood 107 (2006) 3313–3320.